Gyro Error Coefficient Correction Based on GRNN Neural Network

The gyro drift error of cruise missile is the main factor that affects the navigation accuracy. In the non-stop flight segment of cruise missile, the gyro drift error can cause the deviation in the horizontal and vertical direction through affecting the acceleration. Neural networks have a strong self-learning nature, adaptive capacity and non-linear properties. In the paper, we constructed mathematic models and adopted the GRNN neural network to amend the gyro error coefficient of cruise missile really in dynamic flying. The simulation results showed that the method proposed in the paper played well for reducing the gyro drift error, it is also providing a new method to reduce the error of guidance tools.